Moral testimony and moral epistemology

Ethics

Volumn 120, Issue 1, 2009, Pages 94-127

  Hills, Alison ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 73949114196     PISSN: 00141704     EISSN: None     Source Type: Journal    
DOI: 10.1086/648610     Document Type: Article

References (65)

1. By "moral testimony" I mean testimony with explicitly moral content. This might be testimony about what is morally right or wrong: "It is morally right to help those in need." Or it might be testimony using "thick" ethical terms, such as: "It is courageous to face danger without flinching." Or it might be about our moral reasons for action: "I have moral reason not to lie." I will not try to give a definition of what it is for the content of a proposition to be explicitly moral, but I will assume that we can, at least roughly, distinguish between these propositions and others that have what I will call ordinary nonmoral factual content. Of course there is also a controversy about ordinary testimony, namely whether testimony is a basic source of justification (" nonreductionism") or not ("reductionism"), discussed, for example, in Jennifer Lackey and Ernie Sosa, eds., The Epistemology of Testimony (Oxford: Oxford University Press, 2006). But this debate can be put to one side. If Eleanor was learning ordinary nonmoral facts from her friend, both reductionists and nonreductionists would regard her as justified (for she knows that her friend is trustworthy and reliable).
2. Those who think we can gain knowledge from moral testimony include Karen Jones, "Second-hand Moral Knowledge," Journal of Philosophy 96 (1999): 55-78;
3. Julia Driver, "Autonomy and the Asymmetry Problem for Moral Expertise," Philosophical Studies 128 (2006): 619-44;
4. and Robert Hopkins, "What Is Wrong with Moral Testimony?" Philosophy and Phenomenological Research 74 (2007): 611-34.
5. Elizabeth Fricker, in her "Testimony and Epistemic Autonomy," in Lackey and Sosa, Epistemology of Testimony, is also optimistic.
6. Those who disagree include R. P. Wolff, "The Conflict between Authority and Autonomy," in Authority, ed. Joseph Raz (Oxford: Blackwell, 1990). The terms "optimist" and "pessimist" are from Hopkins, "What Is Wrong with Moral Testimony."
7. Bernard Williams, Making Sense of Humanity (Cambridge: Cambridge University Press, 1995), 205.
8. However, as Hopkins points out, some moral truths are trivial, and some nonmoral matters (Will the rope hold my weight? Will the bridge collapse if I stand on it?) are extremely important. So this cannot explain why, in general, deferring to experts and trusting testimony is more acceptable for ordinary nonmoral matters than for moral questions (see Hopkins, "What Is Wrong with Moral Testimony?" 621-23).
9. Hopkins makes this argument for gaining moral knowledge by testimony (ibid., 623-26).
10. Note that the assumption that there is moral knowledge is compatible with moral realism, the theory that there are truths in ethics that are not dependent on our beliefs and attitudes about them. But it is also compatible, I believe, with a wide range of other antirealist theories, including sophisticated forms of subjectivism and noncognitivism, quasi-realism, and moral fictionalism (though they may have to give a revisionary account of what moral knowledge is).
11. The defense is given by Jones, "Second-hand Moral Knowledge";
12. Driver, "Autonomy and the Asymmetry Problem for Moral Expertise";
13. and Hopkins, "What Is Wrong with Moral Testimony?"
14. Hopkins ("What Is Wrong with Moral Testimony?") addresses the puzzle about moral testimony and suggests that the best explanation is that though moral testimony can give us knowledge, there are reasons (perhaps moral reasons) why we ought not to use it. In some ways, this article is an elaboration of that suggestion.
15. It is important to note that this conception of moral understanding-understanding why p (where p is some explicitly moral proposition)-differs from some other types of understanding. It is not the same as understanding the proposition itself, for example, understanding what "it is morally right to help those in need" means. I assume that you need to understand the proposition itself to believe the proposition, let alone to know it or to understand why it is the case. My conception of understanding why p (where p is some moral proposition) should also not be confused with the idea of understanding morality as a whole. I do not have anything to say here about what it is to understand morality (considered as a whole subject area), though it may include having some moral knowledge or some moral understanding in my sense. My conception of moral understanding should also be distinguished from understanding that p (even where p is a moral proposition), a phrase that I think is often used as equivalent to believing that p (or perhaps knowing that p). It is also important to note that I am giving an account here specifically of moral understanding, not of understanding generally.
16. Other philosophers who have written about understanding in general (rather than moral understanding specifically) have taken a different view. For example, Linda Zagzebski, "Recovering Understanding," in Knowledge, Truth and Duty: Essays on Epistemic Justification, Responsibility and Duty, ed. M. Steup (Oxford: Oxford University Press, 2001), develops a conception of understanding according to which understanding is transparent-if you understand that p, you understand that you understand that p-and which is not factive: you can understand that p even though p is false. This is completely different from the concept of moral understanding in which I am interested, and it is also, I think, rather different from our everyday use of the term.
17. Jon Kvanvig, The Value of Knowledge and the Pursuit of Understanding (Cambridge: Cambridge University Press, 2003), has a concept of understanding more similar to mine (though my account of moral understanding is developed in ways with which he might not agree, especially with regard to the importance of certain kinds of cognitive ability). I would like to emphasize that moral understanding as described here is compatible with a variety of metaethical theories, including moral realism, but also with various forms of antirealism (including sophisticated subjectivism, quasi-realism, and fictionalism), provided that they include a conception of truth in ethics.
18. For example, Philip Kitcher, "Scientific Knowledge," in Oxford Handbook of Epistemology, ed. P. Moser (Oxford: Oxford University Press, 2002);
19. James Woodward, Making Things Happen: A Theory of Causal Explanation (Oxford: Oxford University Press, 2003), 179;
20. Peter Lipton, Inference to the Best Explanation, 2nd ed. (London: Routledge, 2004), 30.
21. Stephen Grimm ("Is Understanding a Species of Knowledge," British Journal for the Philosophy of Science 57 [2006]: 515-35) defends a view of understanding according to which it is a "species of knowledge" that involves a "psychologically thick" notion of belief: if you understand that p, you must have "grasped" that p. He does not discuss issues such as accepting that p on the basis of testimony, so it is not clear whether or not he would agree that understanding differs from knowledge in that regard.
22. This testimony might be the trigger that Eleanor needs to think through the matter herself. She might, as a result, grasp why the welfare of animals is important, in such a way that she could make similar judgments about similar cases. If so, then she would understand why p.
23. A very similar case is presented by Duncan Pritchard, "Knowledge,Understanding and Epistemic Value," in Epistemology, ed. Anthony O'Hear (Cambridge: Cambridge University Press, forthcoming), and he draws the same conclusion, that understanding differs from knowledge that and knowledge why.
24. Note that moral understanding does not require a grasp of reasons "all the way down": this would be far too demanding. Nor does it require a grasp of philosophically sophisticated justifications or of moral theory. It does not need any more than the kind of grasp of reasons that most of us have when we realize that we should help others, take care of our friends and family, and so on.
25. Consider the different grounds that you might have for believing a moral proposition p, for example "Action X is wrong." You might base your belief that action X is wrong on the grounds that X is a lie and that lying to others is wrong because it fails to respect them, for example. These factors are reasons why it is wrong, that you could use to explain why that action is wrong. By contrast, you may have evidence that is not of the reasons that make the proposition true, that could not be used to explain why it is true. One obvious source of such evidence is testimony. If a reliable person tells you that p, or even if you discover that they believe that p, this is evidence for p, but it does not give you a clue why p is the case. You cannot offer as an explanation of why action A is wrong: my friend told me it was. We can call this type of evidence, that is not of reasons why p is true, nonexplanatory evidence. The argument here about the limits of testimonial evidence to moral understanding generalizes, I think, to other forms of nonexplanatory evidence (such as the evidence of whether or not other people believe that p, including whether one's "epistemic peers" believe that p) and has significant implications for the appropriate response to moral disagreements, for example. I discuss these implications further in my forthcoming book, The Beloved Self (Oxford: Oxford University Press).
26. This is perhaps one of the reasons why it seemed plausible to some that the virtues are unified in the sense that if you have one virtue, you must have them all. It is not possible properly to understand why one isolated moral proposition is true. But at the same time, you do not have to grasp the nature and significance of all moral reasons in order to understand one of them.
27. I do not mean to suggest here that moral thought has to take place in isolation. Thinking about moral questions with other people can be a way of coming to moral understanding. But if each individual is to have moral understanding, each must grasp the connections between p and the reasons why p, even if they have acquired that grasp through collective deliberation.
28. I leave it open whether having moral understanding is identical with having these abilities, or whether they are connected in some other way that is not identity-but where having moral understanding requires that you have these abilities (to some extent, at least).
29. These abilities are, I think, individually necessary for moral understanding, and I suspect that they may be jointly sufficient, provided that it is true that p and that q is why p (though I am open to the possibility that other abilities may be required in addition). Understanding is often associated with certain sorts of feeling: a flash of enlightenment; a light dawning. But these are not necessary: you need have no particular feelings at all when you finally come to understand why X is morally wrong. Nor are they sufficient. However much you feel the light dawning as you confidently turn aside from the needy, you do not understand-you cannot understand-that it is right to do so.
30. For example, Elizabeth Anscombe argued that dropping the atomic bomb on Hiroshima was wrong because intentionally killing innocent civilians in war is absolutely wrong ("Mr. Truman's Degree," in The Collected Philosophical Papers of G. E. M. Anscombe, vol. 3, Ethics, Religion and Politics [Oxford: Blackwell, 1981]). Let us assume that she was mistaken about that: intentionally killing innocent civilians in wartime can be permissible provided that the benefits are sufficiently great. Let us also assume that in Hiroshima, the benefits were not sufficiently great. Given these (controversial) assumptions, Anscombe was right to condemn dropping the bomb, and she partially understood why doing so was wrong: that it involved the killing of civilians was certainly relevant. But she did not fully understand why it was wrong, though she believed that she did.
31. This is the same sort of case as presented by Pritchard ("Knowledge, Understanding and Epistemic Value") for a similar conclusion. Pritchard concludes that understanding, but not knowledge, is compatible with "environmental" epistemic luck, though he denies that it is compatible with all forms of epistemic luck. I am inclined to go further and to say that even if your own textbook was extremely inaccurate, but that you read in it the truth that Stalin killed millions, you could as a result understand why he was an evil person, though obviously you could not know that he was evil or know why on that basis (I think that Pritchard would deny this). I am inclined to think that you have moral understanding because your belief that Stalin killed millions is true and you formed it reasonably (assuming that you had no reason to think that your textbook was generally inaccurate) and you have grasped correctly he was evil because he killed millions-and this is all that is required for understanding why he was evil.
32. The best known defender of the claim that know-how is not propositional knowledge is Gilbert Ryle, "Knowing How and Knowing That," in Gilbert Ryle: Collected Papers, vol. 2 (New York: Barnes & Noble, 1971), 212-25.
33. The thesis is also endorsed by David Lewis, "What Experience Teaches," in Mind and Cognition: A Reader, ed. W. G. Lycan (Oxford: Blackwell, 1990), 499-519.
34. Recently, the opposite claim-that know-how is a form of propositional knowledge-has been forcefully argued by Jason Stanley and Tim Williamson, "Knowing How," Journal of Philosophy 98 (2001): 411-44.
35. The argument given by Stanley and Williamson is complicated, and I cannot do justice to it here. They claim that knowing how is a species of propositional knowledge, specifically that, for example, "Hannah knows how to ride a bicycle" is true relative to a context c "if and only if there is some contextually relevant way w such that Hannah stands in the knowledge-that relation to the Russellian proposition that w is a way for Hannah to ride a bicycle, and Hannah entertains this proposition under a practical mode of presentation" (ibid., 430).
36. Stanley and Williamson ("Knowing How") press the point that having knowledge how can involve having certain complex dispositions, and yet knowledge how may still be a species of propositional knowledge, so they would not find convincing the first reason I give for thinking that moral understanding and propositional knowledge are distinct. But even if they are right about that, it is still true that moral understanding seems to be compatible with epistemic luck in a way that propositional knowledge is not, in which case understanding why p (where p is a moral proposition) cannot be a species of propositional knowledge.
37. T. M. Scanlon, What We Owe to Each Other (Cambridge, MA: Harvard University Press, 1998), 147-60. According to Scanlon, the justification of your action must be one that others could not reasonably reject.
38. It may be very unlikely that anyone will ever ask you to justify what you do. You may know that no one will ask you: perhaps you acted in secret; maybe everyone is more concerned with their own affairs. Why should it be important that you can justify yourself to them, if you know you will never have to? Perhaps you find yourself imagining other people watching you, and you imagine trying to justify yourself to them. You might feel very uncomfortable if you found yourself unable to explain to them in your own words why what you were doing was right. But if you preferred not to exercise your imagination in this way, it is difficult to see why you should care about justification to others at all, in situations where no one wants a justification.
39. Compare Aristotle: "the "product" is brought to completion by virtue of a person's having wisdom and excellence of character, for excellence makes the goal correct, while wisdom makes what leads to it correct" (Nicomachean Ethics, trans. Sarah Broadie and Christopher Rowe [Oxford: Oxford University Press, 2002], 1144a6-1144a9). I want, as far as possible, to leave open the connection between the two-for example, I do not intend to take a stand on whether, if you have good judgment, you need a separate motivation in order to act-or whether you must have a good motivation, if your judgment is good.
40. John McDowell defends a similar view, though he says that the virtuous have a "reliable sensitivity to a certain sort of requirement which situations impose on behaviour" ("Virtue and Reason," in Virtue Ethics, ed. Roger Crisp and Michael Slote Oxford: Oxford University Press, 1997;
41. first published in The Monist 62 (1979): 331-50, 142. He regards this sensitivity as delivering knowledge, though as I have explained, I think it is better characterized as moral understanding. There are other important features of McDowell's conception of virtue on which I do not want to make a stand (for example, his view that this "knowledge" is sufficient for action).
42. In order to have some virtues, you may not need to form explicitly moral beliefs. It may be sufficient to be generous, for example, that you simply judge that someone else is in need and that you could help them, and you go on to do so. But it is likely that you will have to make explicitly moral judgments some of the time, and with regard to some virtues-like justice, for example-you will almost certainly need to judge outcomes in moral terms (in this case, as just or unjust). There is in fact no sharp distinction between moral and nonmoral motivations, and moral and nonmoral judgments. For example, one of the important ways that we recognize moral reasons is through emotional responses, some of which involve a distinctively moral attitude, or a combination of nonmoral response and moral judgment: we may feel anger at injustice or at cruelty; admiration at an example of courage; and anxiety at whether we can live up to the same standards-respect for others can be a feeling which is not explicitly moral (which does not involve explicitly moral beliefs) or a distinctively moral attitude toward others that we deliberately take up. There is often continuity between our nonmoral attitudes and emotions and those that involve explicitly moral descriptions and lead us to form explicitly moral beliefs. An honest person, for example, grasps why honest actions are important and why she ought to perform them. A courageous person is motivated to face danger precisely when she grasps that it is worth doing so. Much (though not all) of the motivations and judgments associated with the virtues are explicitly moral.
43. For example, Claire's belief that A is right is not counterfactually dependent on R or her belief that R, but on what Judith says. Of course, it will be difficult to distinguish these different dependences in practice, since in fact what Judith says about whether A is right will itself depend on R. But though Judith is virtuous and trustworthy, she is not infallible, and Claire's belief depends on Judith's testimony, not on R (or her belief that R), when the two come apart. Counterfactual dependence is only a sign of the basis of one's belief though, and it is not an infallible sign. If Claire gains and uses moral understanding, for instance, her belief that A is right may be based both on Judith's testimony and on her belief that R; both of these explain why she has that belief. I take it that it is sufficient for proper orientation that one explanation for your moral belief is your moral understanding; it may have other explanations too. When Claire does not have or does not use moral understanding, by contrast, her belief that A is right is explained by Judith's testimony, not by her belief that R.
44. As Aristotle says, the virtuous person has the ability to "see what is true in every set of circumstances, being like a carpenter's rule or measure for them" (Nicomachean Ethics, 1113a32-1113a34).
45. Immanuel Kant, Groundwork of a Metaphysic of Morals, trans. H. J. Paton London: Routledge, 1991;
46. page references cite the volume and page number of Kants gesammelte Schriften, published by the Preussische Akademie der Wissenschaften [Berlin: W. de Gruyter, 1902]), 4:397. What is it to act for the right reasons? Kant himself thought that morally worthy action was doing your duty from duty-though it is not exactly clear what that means. One possibility is that to do the morally right act for the right reasons, you must choose it under an explicitly moral description: "the morally right action" or "the action which I have overriding moral reason to perform" or "my duty." But that is a mistake. The use of the concept of a duty or of moral rightness-or indeed the use of any explicitly moral concept-is not essential to morally worthy action. What matters is that you respond to moral reasons, not that you do so under an explicitly moral description. In fact, as Williams emphasizes, we sometimes expect people to respond to moral reasons without thinking of them in explicitly moral terms, and responding to them under an explicitly moral description, "it was his wife; and that in situations of this kind it is permissible to save one's wife," seems to be "one thought too many"
47. Bernard Williams, Moral Luck [Cambridge: Cambridge University Press, 1981], 18). Of course, Williams's argument is compatible with moral beliefs playing an important background role; for example, your recognition that it is morally acceptable to save your wife might influence what you do even though it does not consciously cross your mind. But it is likely that at times when you act, you will form some moral beliefs and they will play an important role. Sometimes there is no clear distinction between explicitly moral and nonmoral considerations. Sometimes nonmoral motivations lead you astray and you need to form moral beliefs to correct them. Sometimes what is morally required is complicated and you need to deliberate and think in explicitly moral terms in order to work out what is morally right in your circumstances. It does not follow, of course, that our forming explicitly moral beliefs does actually help us to respond to moral reasons. Our moral beliefs can fail to fulfill this function in (at least) two ways: first, they may be false, and lead us astray; second, we may fail to act on them. And of course, they can fail in both ways at once. But if you do act on your moral beliefs, I will argue, you need to have and to use moral understanding in order to act well.
48. Arpaly's final account of morally worthy action involves sensitivity to the rightmaking features of an action-plus deep moral concern-which is similar to my conception of the two aspects of orientation (Nomy Arpaly, Unprincipled Virtue [Oxford: Oxford University Press, 2003], 84).
49. Ibid., 74. Arpaly describes Ron as believing deeply that killing a person is not generally immoral but that killing a fellow Jew is a grave sin.
50. Ibid.
51. Ibid.
52. Of course, if he believed that killing Tamara was wrong on the basis of a false belief, he would not have knowledge. But he does not believe it on that basis; he believes it because he was told it by a source he knows to be trustworthy, and it is hard to see that his justification for that belief is undermined because he has independently come up with a mistaken explanation for it. Compare: you have a lot of knowledge from perception, and sometimes you come up with mistaken explanations for it; for example, you know that the sun rose in the morning because you saw it, even though you mistakenly believe that it did so because the sun orbits the earth. The mistaken explanation does not undermine your perceptual knowledge.
53. He would then be responding to moral reasons de re (rather than under an explicitly moral description).
54. Ron's action is therefore counterfactually dependent on the rabbi's testimony (specifically, his testimony that the action is right) together with his belief that the rabbi is reliable and trustworthy, but not counterfactually dependent on his belief that Tamara is a person. Of course, this is complicated by the fact that the rabbi also testifies that Tamara is a person and that this fact is morally relevant, but if the rabbi had given a different explanation of the rightness of the action, Ron would still have done what the rabbi told him was right, whatever he believed about Tamara's being a person. This is, I think, grounds for claiming that Ron's belief that Tamara is a person is not among the causes of his action and therefore is not among his reasons for action. It is part of what is sometimes called the "standard model" of acting for reasons, defended, for example, by Donald Davidson ("Actions, Reasons and Causes," in his Essays on Actions and Events [Oxford: Oxford University Press, 1980], 3-19, esp. 9), that acting for the reason R is being motivated to act by the belief that R. The standard model has recently come under attack on the grounds that your action can be caused by a belief and a desire without your acting for reasons
55. (David Velleman, The Possibility of Practical Reason [Oxford: Oxford University Press, 2000], 2-9). I am sympathetic to these concerns, but I think that a causal connection between your belief that R and your action is going to be part of any plausible account of acting for the reason R, even if it is not sufficient: in order for a reason for an action to be your reason for action, it must cause your action.
56. Kieran Setiya (Reasons without Rationalism [Princeton, NJ: Princeton University Press, 2007], 28-48), however, suggests that though this is true in the "overwhelmingly typical" case, it may not be true in unusual cases, such as in self-deceptive action. He suggests a different connection: "In what we may think of as the paradigm case of acting for a reason, an agent takes it that he is hereby doing J because he believes that p, where this is the 'because' of motivation"
57. (ibid., 45). In other words, to act for the reason p, you must believe that your belief that p caused you to act (even if in fact it did not). Circumstances in which the belief that p did not in fact cause you to act (though you think that it did) Setiya describes as "marginal" cases of acting for a reason. But I think a better description of this kind of case is: you thought that you acted for the reason that p but in fact you did not-since your belief that p did not in fact cause you to act. In any case, Ron has no reason to think that his belief that Tamara is a person caused him to act (either directly or indirectly, via a belief that it is right not to kill her). So even if Setiya's account of acting for reasons is correct, Ron has no grounds for claiming that he is acting for the right reasons (and if my interpretation of what Setiya calls the "marginal" case is correct, he does not act for the right reasons).
58. Arpaly argues that the moral worth of an action corresponds to the moral praiseworthiness or blameworthiness of the agent, that is, the extent to which she deserves praise or blame for performing those actions (Unprincipled Virtue, 69). If this is right, there is another way in which focusing on morally worthy action could be self-indulgent: the agent could primarily be motivated to be praised. It would not, of course, be particularly admirable to aim to perform morally worthy actions because you want praise (though of course Arpaly does not suggest that it is). But according to my conception of acting well, morally worthy actions are right actions performed for the right reasons. Since to get praise is not usually the right reason for action, someone who was motivated primarily by praise would not be performing morally worthy actions. Instead, they would choose right because there were forceful moral reasons to do so. Acting for these moral reasons is not selfserving or self-indulgent.
59. Or, more precisely, it is essential to morally worthy action that is based on explicitly moral belief. This is, I think, similar to Aristotle's view that the end of phronesis is not the production of something distinct from the productive process but "doing well itself serves as end" (Nicomachean Ethics, 1140b7-1140b8).
60. Children may acquire moral understanding by learning the habit of doing what is right and not what is wrong, and eventually coming to grasp the reasons why some actions are morally right and others wrong (this is a simplified version of Aristotle's conception of moral education). At the first stage of this acquisition, they learn from the testimony and example of those around them what they have moral reason to do (and not to do)-so moral testimony plays a crucial initial role-then they habitually do the right thing, and testimony plays less of a role. The final stage of acquiring moral understanding, however, where they come to grasp why it is right to do what they already habitually do, requires them to exercise their own judgment. It is this stage that needs practice, and moral testimony can play at most a guiding role.
61. Driver, "Autonomy and the Asymmetry Problem for Moral Expertise," 639.
62. Someone might reasonably think that moral understanding is valuable but that since it is unavailable to many people (perhaps even to most of us), we should trust the testimony and defer to the judgments of the experts. Plato seems to express this sort of view in the Republic. The Guardians who have undergone the right education and training understand why justice and the other virtues are important, and are just and virtuous themselves. Ordinary people are not capable of going through the education required of the Guardians and so can never have moral understanding. It is best, Plato thinks, if society is set up with the Guardians in charge and with everyone else doing what they are told. The moral understanding that the Guardians have is clearly extremely valuable, on Plato's view, but it is simply not open to all. Deferring to the Guardians is appropriate, not because moral understanding is not worth having, but because, however long and hard they tried, most people would never attain it. The next best option is for them to do what they are told by people who do. (It is, of course, very controversial whether Plato thought that the Guardians had moral knowledge or what I would call moral understanding. For an argument that they acquire knowledge, see Gail Fine, "Knowledge and Belief in Republic V-VII," in her Plato on Knowledge and Forms [Oxford: Oxford University Press, 2003]. For a reading more similar to mine,
63. see Julia Annas, An Introduction to Plato's "Republic" [Oxford: Oxford University Press, 1981]).
64. This is the best interpretation, I think, of a well-known example from Karen Jones, of a situation in which Peter and his housemates are trying to choose someone else to share their house: "White women and women of color had advocated rejecting three white men on the basis of their perceived sexism, and in one case, racism. Peter had a settled and serious commitment to the elimination of racism and sexism, but he was not very good at picking out instances of sexism and racism. Thus, his commitment often remained theoretical and he saw fewer occasions as calling for antisexist and antiracist responses than he might have seen. Such blindness can sometimes indicate insincerity but in Peter's case it did not. He genuinely did want to understand the perspective of the women and he wanted to be able to share it, if indeed it turned out to be correct. He could pick out egregious instances of sexism and racism, and could sometimes see that 'sexist' or 'racist' applied to more subtle instances when the reasons for their application was explained to him, but he seemed bad at working out how to go on to apply the word to nonegregious new cases. Problems arose for Peter when he could not see the reasons why the women were calling someone sexist, and could not see, or could not see as evidence, the considerations that the women thought supported viewing the would-be members as sexist. . . . I think that in this case, Peter should have been willing to accept the women's testimony that these men were sexist. But for him to have done so would have been to accept and act on a moral judgment on someone else's say-so. It would have been to borrow moral knowledge" ( Jones, "Second-hand Moral Knowledge," 59-60). Peter has evidence that he has poor judgment in this area and that despite his best efforts he has been unable to acquire moral understanding. He also has evidence that in trusting the women's testimony he would acquire moral knowledge. In this situation, it may be right for him to do so.
65. I defend the claim that epistemic rationality for moral beliefs is defined in terms of moral understanding in my The Beloved Self.